The Need for Physical Measurements in the Execution and Evaluation of Cloud Seeding Experiments to Enhance Precipitation in Convective Clouds

The large- and meso-scale dynamics determining the characteristics of convective cloud systems down to the small-scale microphysics determining the nucleation and growth characteristics of water droplets and ice particles all form part of the chain of events of precipitation development. Although our knowledge of the individual aspects in the chain has significantly increased in the past twenty years there still exist major gaps about certain physical processes. The efficiency with which clouds produce rain at the surface varies greatly. Precipitation efficiency, defined as the ratio of the rate of rain reaching the ground to the flux of water vapor passing through cloud base, can range from zero in non-precipitating clouds to greater than unity for short times, in very intense, time-dependent, convective systems.

It is clear that aerosols contribute to the observed differences in cloud droplet size distributions between maritime and continental and between non-polluted and polluted convection. In addition, other factors such as cloud base temperature, boundary layer depth and turbulence, thermodynamic profile (moisture and parcel buoyancy) that vary between land and ocean regions and from day to day in a certain region, could also be contributing to observed differences in precipitation processes or acting in concert with aerosol effects. In addition, the initial cloud droplet spectra at cloud base to a large extent determines the microphysical processes of precipitation formation (water and ice) at higher levels in the clouds and how these processes are modified via cloud seeding with droplet and ice nucleating aerosols.

Variations in meteorological conditions can dominate the effects of seeding and are often times much larger than the effect of seeding (10-100 times). These variations can occur in space and in time and can significantly affect the results from any cloud seeding experiment on a day to day basis as is evident from the analyses of randomized seeding experiments in the past. Where historical randomized cloud seeding experiments have primarily depended on a single statistical test assuming that the samples are randomly drawn from the same distribution of potential values (treatment application for these measurements was at random) and many of these experiments were inconclusive. This has major impacts on the execution and evaluation of both operational and randomized cloud seeding experiments.

The presentation will provide an overview of methods and analyses from recent experiments to highlight some of the issues related to natural variability in the historical statistical analyses and evaluation of rainfall enhancement experiments. It will also provide a future framework for conducting operational and randomized cloud seeding experiments.